Start with outcomes
Agree the service, environmental, cultural and affordability outcomes with those affected. Translate them into evaluation criteria before selecting preferred assets. Keep statutory obligations distinct from discretionary benefits.
Compare credible alternatives
Include a defined baseline and demand-side or operational responses where feasible. Assess capital and operating costs, delivery constraints, energy use, carbon, reliability and distribution of benefits. Document how qualitative criteria are scored and who set their weights.
Use scenarios to expose fragility
An option that wins under one forecast can perform poorly if growth is delayed, energy costs rise or source yield changes. Test plausible combinations and identify common early actions. Adaptive pathways reserve later decisions for when monitoring provides better information.
Make the plan implementable
- Set observable trigger conditions rather than relying only on calendar dates.
- Allow time for investigation, engagement, approvals and construction.
- Assign an owner to each monitoring indicator and action.
- Record when the plan will be reviewed and what evidence could change it.
Define the decision and the alternatives
Start with the service problem, its affected community and the period over which a response is needed. A project objective framed as “build a new reservoir” has already selected a solution. An objective framed around a demonstrated service deficit allows storage, transfer, operations, demand management and source options to be compared.
Establish the baseline and the consequences of doing the minimum. Include committed works and realistic maintenance rather than comparing every option with an artificially deteriorating system. Identify dependencies such as land, power, approvals, workforce and upstream treatment.
Compare whole-of-life quantities consistently
Use a common assessment period, price basis, service outcome and treatment of residual value. Include capital, operation, energy, maintenance, renewal and decommissioning where material. Document the discount rate and escalation assumptions adopted for the project. The Atlas does not prescribe an economic rate or approval method.
Separate an option’s nominal capacity from the service it reliably delivers under the assessed scenarios. A cheaper asset can be poor value if another bottleneck prevents its use. Equally, an oversized early investment can impose avoidable carrying and operating costs when growth arrives later than expected.
Use multi-criteria analysis transparently
Criteria can represent service, cost, environmental effects, social and cultural outcomes, constructability and flexibility. Define each criterion so reviewers can understand what a score means. Avoid counting the same benefit several times under different names. Keep evidence quality and uncertainty visible.
Weights express preferences, not scientific facts. Test whether plausible changes in weights or uncertain scores alter the preferred option. If the ranking is fragile, the decision may need further investigation or a staged commitment. A weighted total should not hide an unacceptable constraint that an option must satisfy regardless of score.
Design an adaptive pathway
| Element | What to specify |
|---|---|
| Near-term action | A useful step supported across several plausible futures. |
| Signpost | An observable variable such as verified connected demand or dependable source output. |
| Trigger | The condition that starts a decision or delivery action. |
| Lead time | Design, approval, procurement and construction time with contingency. |
| Next option | A feasible response whose corridor, land or other dependencies remain available. |
A trigger set at the point of service failure is too late for an asset that takes years to deliver. Work backwards from the projected constraint and delivery time, then test uncertainty. Preserve strategic options through corridor or site planning where justified, while avoiding premature construction.
Link model evidence to an accountable recommendation
The recommendation should explain the problem, the compared options, the governing scenarios, the evidence supporting performance and the residual uncertainty. Distinguish a modelled result from an assumption and a stakeholder preference. State what could change the decision and how it will be reviewed.
Bring operators, asset managers, communities and Traditional Owners into the relevant stages. Their knowledge can change feasibility, failure consequences and the definition of a successful outcome. Keep the assessment record useful for future reviewers who were not present during the original study.
Follow the water-security framework for uncertainty across supply sequences and the network-planning guide for local capacity decisions. The same traceability should connect both scales.
Sources & further reading
- Water resources planning guideline ↗Environment Agency, Natural Resources Wales & Ofwat · Living guidance
External source · Checked 24 September 2026 - Best Practice Modelling Guidelines ↗eWater · 2011
External source · Checked 24 September 2026
Source findings are distinguished from editorial interpretation. Apply current local criteria and project evidence when making engineering decisions.